Acousto-mechanical and thermal properties of clotted blood

a)
Volodymyr M. Nahirnyak
Department of Physics, University of Cincinnati, Cincinnati Ohio 45221-0011
Suk Wang Yoon
b
and Christy K. Holland
c
Department of Biomedical Engineering, University of Cincinnati, Medical Science Building, Room 6167,
231 Albert Sabin Way, Cincinnati, Ohio 45267-0586
Received 28 November 2005; revised 7 April 2006; accepted 10 April 2006
The efcacy of ultrasound-assisted thrombolysis as an adjunct treatment of ischemic stroke is being
widely investigated. To determine the role of ultrasound hyperthermia in the process of blood clot
disruption, the acousto-mechanical and thermal properties of clotted blood were measured in vitro,
namely, density, speed of sound, frequency-dependent attenuation, specic heat, and thermal
conductivity. The amplitude coefcient of attenuation of the clots was determined for 120 kHz,
1.0 MHz, and 3.5 MHz ultrasound at room temperature 202 C. The attenuation coefcient
ranged from 0.10 to 0.30 Np/ cm in porcine clots and from 0.09 to 0.23 Np/ cm in human clots. The
experimentally determined values of specic heat and thermal conductivity for porcine clotted blood
are 3.20.5 10
3
J/ kg K and 0.550.13 W/ m K, respectively, and for human clotted blood are
3.50.8 10
3
J/ kg K and 0.590.11 W/ m K, respectively. Measurements of the
acousto-mechanical and thermal properties of clotted blood can be helpful in theoretical modeling
of ultrasound hyperthermia in ultrasound-assisted thrombolysis and other high-intensity focused
ultrasound applications. 2006 Acoustical Society of America. DOI: 10.1121/1.2201251
PACS numbers: 43.35.Wa, 43.80.Ev CCC Pages: 37663772
I. INTRODUCTION
An estimated 700 000 people experience a stroke in the
U.S. every year, resulting in an aggregate lifetime cost of
$56.8 billion AHA, 2005. The only therapy for ischemic
stroke that is approved by the FDA is the thrombolytic agent,
recombinant tissue plasminogen activator rt-PA. Recombi-
nant t-PA is moderately effective in lysing thrombi in is-
chemic stroke patients and it improves neurologic decits if
given within three hours after the onset of stroke symptoms
Wolpert et al., 1993. Unfortunately, thrombolytics also can
cause intracerebral hemorrhage. Thus adjuvant therapies that
lower the rt-PA dose or increase its efcacy would represent
a signicant breakthrough Polak, 2004. Experimental evi-
dence has demonstrated that ultrasound used as an adjuvant
to rt-PA can increase thrombus dissolution in an in vitro
model Shaw et al., 2001a, b, 2006; Holland et al., 2002.
Mild heating of only a few degrees and mechanical effects,
such as acoustic streaming, radiation force, cavitation, and
accompanying microstreaming all could contribute to the en-
hanced thrombolysis.
Siddiqi et al. 1995 reported increased permeation of
brinolytic enzymes into a brin gel due to pulsed ultra-
sound exposure at 1 MHz with an intensity of 2 W/ cm
2
and
50% duty cycle. The authors attributed the enhanced perme-
ability to ultrasound-induced cavitation. Degassing the uid
in which the brin gel sample was immersed signicantly
reduced the effect of brinolysis.
Increased penetration of radiolabeled rt-PA into clots
during 1-MHz insonation at 4 W/ cm
2
was observed by
Francis et al. 1995; Francis, 2001. Ultrasound exposure
increased the transport of rt-PA into the plasma clot resulting
in deeper penetration. Cavitation has also been shown to be
an important physical mechanism in the ultrasound-assisted
thrombolysis Everbach and Francis, 2000. Cavitation is
thought to loosen the brin network, increase the penetration
rate of the brinolytic agent, and create additional binding
sites causing the fragmentation of the thrombus. However,
other mechanisms were hypothesized to participate in the
enhancement because the application of overpressure, which
prevents the nucleation of inertial cavitation, did not com-
pletely remove the acceleration of thrombolysis.
Sakharov et al. 2000 provided evidence for streaming
and heating during continuous wave ultrasound-enhanced ly-
sis of plasma clots at 1 MHz. A temperature increase in the
clot of a few degrees produced a twofold increase in lytic
rate. This group attributed 30% of the effect of ultrasound-
enhanced brinolysis to heating and the remaining 70% to
microstreaming. The peak-to-peak pressure employed was
0.26 MPa and the corresponding time-averaged intensity was
2.3 W/ cm
2
. Enhanced thrombolysis was also achieved by
heating and mild stirring without ultrasound exposure.
There is general agreement that a temperature increase is
only minimally responsible for accelerated thrombolysis
Dick et al., 1998; Francis et al., 1992; Lauer et al., 1992;
Blinc et al., 1993; Harpaz et al., 1993; Olsson et al., 1994;
Sakharov et al., 2000.
a
A portion of this work was presented in Acousto-mechanical and thermal
properties of clotted blood, at the 149th meeting of the Acoustical Society
of America, Vancouver, Canada, May 2005.
b
Dr. Suk Wang Yoon is a visiting professor from SungKyunKwan Univer-
sity, Suwon 440-746, Republic of Korea.
c
Author to whom correspondence should be addressed. Electronic mail:
christy.holland@uc.edu
3766 J. Acoust. Soc. Am. 119 6, June 2006 2006 Acoustical Society of America 0001-4966/2006/1196/3766/7/$22.50
Suchkova et al. 2002 explored ultrasound-enhanced -
brinolysis at low kilohertz frequencies 27100 kHz in an
attempt to minimize ultrasonic heating and concomitant ad-
verse bioeffects. Signicant enhancement of brinolysis was
achieved, with the greatest effect observed at 27 kHz. The
largest effect was observed with continuous-wave ultra-
sound. A signicant acceleration of ultrasound fybrinolytic
effect was also observed with peak intensity of 1 W/ cm
2
and
duty cycles of 10% and 1%.
According to Cintas et al. 2004, the use of galactose-
based microbubbles increased the ultrasound-assisted brin-
olytic effect during low- intensity, 2-MHz ultrasound expo-
sure of whole human blood clots with rt-PA. After a bolus of
a 0.2 mL 400 mg/ ml of D-Galactose Palmitic Acid Levo-
vist was added to the closed-loop ow system containing
a clot, the authors reported a 31% clot mass loss after 30
-min insonication at an intensity of 0.46 W/ cm
2
in the
presence of rt-PA. Clots exposed to rt-PA alone exhibit only
a 13.2% mass loss. These authors concluded that cavitation
and microstreaming were the main mechanisms responsible
for clot disruption and fragmentation. These effects may al-
low a deeper penetration of rt-PA into the clot which in turn
exposes a larger percentage of the brin network to rt-PA.
An analytical model is being developed to determine the
contribution of mild heating as a mechanism for the accel-
eration of the enzymatic activity of rt-PA during lysis using
pulsed ultrasound over the frequency range of
120 kHz to 3.5 MHz. For such a computational model to be
developed, the acousto-mechanical and thermal properties of
clotted blood must be well characterized. Duck 1990 has
collated an extensive literature on the physical properties of
tissue, but the specic heat and thermal conductivity of clot-
ted blood have not yet been explored. Therefore, experiments
were conducted in vitro with porcine and human clots at
room temperature to measure the density, specic heat, ther-
mal conductivity, speed of sound, and attenuation over the
frequency of 120 kHz to 3.5 MHz.
II. METHODS
A. In vitro clot model
The synergistic thrombolytic effect of rt-PA and 120
-kHz ultrasound was assessed previously in vitro in a porcine
clot model Holland et al., 2002. To determine the thermal
and acousto-mechanical properties of these clots, the same
blood clot production protocol was followed, which was ap-
proved by the local institutional animal care and use com-
mittee. Whole blood clots were prepared by aliquoting
1.5 ml arterial porcine or venous human blood into the 8
-mm inner diameter glass Vacutainer tubes, immersing the
tubes in a 37 C water bath for 3 h and storing the clots at
5 C prior to use in comparative ultrasound and rt-PA stud-
ies, which ensured complete clot retraction. This type of clot
is fairly similar to physiologic venous clots and avoids the
alteration of the clot structure by employing anticoagulated
blood and the addition of thrombin. Additional aliquots of
blood from each pig or human were used to obtain a com-
plete coagulation panel from Antech Diagnostics Chicago,
IL, including D-Dimer, A-PTT, brinogen and prothrombin
time testing, as well as a complete blood count. Most pigs
used as part of this study were found to be slightly anemic,
with hematocrits in the range of 25% to 35%. Only donors
with values in the range 10900 ng/ ml for the D-Dimer test,
1025 s for A-PTT, 250700 mg/ dl for the brinogen con-
centration, and 913 s for prothrombin time were considered
to be acceptable. The resulting clots were normally dark red
in color, roughly cylindrical in shape with an average diam-
eter of 710 mm, and with the typical mass about 0.5 g for
each of them, shown in Fig. 1.
B. Density, calorimetric, and thermal conductivity
measurements
For the measurements of clot density, the volume of the
clots was determined using a uid displacement method with
an uncertainty of 0.08 cm
3
. A standard laboratory scale
E200, Mettler-Toledo, Inc., Columbus, OH, was used to
assess the mass of each clot in physiologic saline at room
temperature 222 C with an uncertainty of 0.01 g. The
overall uncertainty in the density measurements was 0.02
10
3
kg/ m
3
. A total of 202 clots from 28 human subjects
and a total of 336 porcine clots from 38 pigs were employed
for the density and calorimetric measurements. Thus the in-
uence of physiologic variability and experimental uncer-
tainty were captured in the standard deviation of the density
and calorimetric measurements.
Direct calorimetric measurements using calibrated
E-type thermocouples Omega Engineering, Inc., Stanford,
CT were performed to determine the specic heat of human
and porcine clots relative to the specic heat of a standard
uid, physiological saline 0.9% sodium chloride solution.
The experimental setup included two 236 ml Styrofoam
containers with insulating lids, which contained E-type ther-
mocouples. Two digital thermometers Model HH 506-R,
Omega Engineering, Inc., Stanford, CT with thermocouples
attached to them were used to record temperatures within the
containers throughout the experiment. The method of mix-
tures Semat, 1953 was used to determine the heat capacity
of the sample clots in saline held at an inital temperature of
FIG. 1. Photo of a blood clot sample. Whole blood clots were prepared from
either fresh porcine or human blood by aliquoting 1.5 or 2.0 ml into 10
-ml glass tubes BD Vacutainer, Franklin Lakes, NJ, immersing the tubes
in a 37 C water bath for 3 h and storing the clots at 5 C for at least 3 days
prior to assessment of the properties, which ensured complete clot retrac-
tion. The typical size of the clot samples was about 710 mm in diameter
and the typical mass of each clot was about 0.5 g.
J. Acoust. Soc. Am., Vol. 119, No. 6, June 2006 Nahirnyak et al.: Physical properties of clotted blood 3767
61 C and combined with a second aliquot of saline at
room temperature 211 C. The uncertainty in the tem-
perature measurements was 0.06 C and the uncertainty in
the measurements of the clot mass was 0.01 g. The propa-
gated error for the specic heat measurements in this case
was 0.310
3
J/ kg K.
The comparative steady-state method Kreith and Black,
1980 was used for determination of the thermal conductivity
of whole blood clots in the experimental apparatus shown in
Fig. 2. The sample chamber is inserted between two 25
2525 cm
3
Plexiglas water tanks held at two different
temperatures, T
1
and T
2
. A thermal gradient was thus created
horizontally across the cylindrical experimental cell made of
polyethylene-terephthalate glycol PETG containing both
the clot sample and a uid of known thermal conductivity,
namely deionized, degassed water. PETG has a signicantly
higher coefcient of thermal conductivity 0.33 W/ m K
than air 0.024 W/ m K, thus allowing us to simplify the
experimental setup to one-dimensional heat ow along the
axis of the cylindrical cell. Six Luerlock connectors were
attached to the experimental cell as ports for positioning
thermocouples and expunging air four along the chamber
containing clot and two along the chamber containing water.
The diameter of the cylindrical chambers containing the
standard uid and clot sample was 11.2 mm. The length of
the chamber containing the clot sample was 55 mm and the
length of the chamber containing water was 35 mm. A 1.27
-mm-thick copper membrane with high thermal conductivity
separated the clot chamber from the chamber lled with wa-
ter. The membrane was necessary to prevent contamination
of the water from the sample chamber. Hypodermic, 0.29
-mm-diam T-type thermocouples HYP1-30, Omega Engi-
neering, Inc., Stanford, CT were used to monitor the tem-
perature at four different points along the chambers. Two
stainless steel pistons of diameter 11.2 mm and length
80 mm provided thermal contact between the two water
tanks and the sample chambers. This experimental design
allowed the position of the thermocouple tips to be adjusted
within the clot and the temperature drop at different locations
along the clot to be measured. The magnitude of the thermal
gradient between the tanks was approximately 5 C which,
given losses in the stainless steel pistons, corresponded to an
approximate 1 to 1.5 C thermal gradient across the clot
sample.
To validate our method of measuring thermal conductiv-
ity, the sample chamber was lled with a uid of known
thermal conductivity, ethanol, as a test. A time period of an
hour was required to allow the temperature gradients to sta-
bilize in the sample chambers. The coefcient of thermal
conductivity of the investigated substance either clot or eth-
anol was calculated from the following formula:

unknown
=
w
T
2
L
2
L
1
T
1
, 1
where
unknown
is the thermal conductivity of the investi-
gated substance,
w
is the thermal conductivity of water,
T
1
=T
2
T
1
is the magnitude of the thermal gradient
across the sample under investigation, L
1
is the distance
between the two thermocouples in the sample, T
2
=T
4
T
3
is the thermal gradient across the water sample, and
L
2
is the distance between the two thermocouples in the
water-filled chamber. The uncertainties in the measure-
ments of temperature and distance were 0.06 C and
0.2 mm, respectively. The propagated error in determina-
tion of the coefficient of thermal conductivity was
0.02 W/ m K. The thermal conductivity measured using
ethanol as the investigational substance was
0.200.02 W/ m K, which compares favorably with the
published value for the thermal conductivity of ethanol,
0.19 W/ m K Kreith and Black, 1980.
The coefcient of thermal conductivity was determined
in three 10-ml specimens of clots prepared according to the
protocol described above. We recorded the thermal gradients
in the clot samples at three different positions within the
clots. The thermal conductivity of porcine and human clots
was determined by employing Eq. 1, averaging the results
of all nine experiments, and computing the standard devia-
tion of the measurements.
C. Sound speed in clotted blood
We used an insertion pulse-echo method to measure the
speed of sound in clotted blood. A polystyrene cuvette of
4.5 ml capacity and dimensions 12.512.546 mm
3
Cur-
tin Matheson Scientic, Inc. Baltimore, MD was immersed
in a water tank and positioned within the focus of a 3.5
-MHz transducer. Ultrasonic pulses were generated with a
function generator Model 33250, Agilent Technologies, Inc.,
Palo Alto, CA, amplied Model AR 150LA, Ampier Re-
search, Souderton, PA, and monitored with an oscilloscope
LT372, LeCroy Corp., Chestnut Ridge, NY. The water tank
was connected to the water thermostat EX111, Neslab In-
struments, Inc., Newington, NH. Experiments were con-
ducted at two temperatures: 20 C and 37 C. The choice of
the working frequency was determined by two factors. First,
the axial resolution of the ultrasound measurement is dic-
tated by the acoustic wavelength, the shorter the better. On
the other hand, the choice of too high a frequency is limited
by the attenuation in clot and cuvette walls, which were
6 mm apart.
FIG. 2. Experimental apparatus for thermal conductivity measurements in
blood clots.
3768 J. Acoust. Soc. Am., Vol. 119, No. 6, June 2006 Nahirnyak et al.: Physical properties of clotted blood
The time delay between echoes reected from the front
face and far wall of the cuvette was used to determine the
speed of sound at two temperatures, 20 C and 37 C. The
pulse repetition period for all experiments was 10 ms which
corresponds to a PRF of 100 Hz. The three-cycle pulse du-
ration was 0.9 s. First, we measured the time delays be-
tween echoes from the front and back walls of the cuvette
when it was lled with water. By using the sound speed in
water at 20 C and 37 C, 1482 and 1524 m/ s, respectively,
we determined the distance between two walls of the cuvette.
After the water was drained, a blood clot was carefully put
into the same cuvette and the time between echoes from the
cuvette walls was measured. The speed of sound in the clot
was calculated by dividing twice the distance between the
cuvette walls by the measured time delay. Three human clot
samples and three porcine clot samples were used and the
mean and standard deviation were calculated. For validation
of the technique, the cuvette was lled with ethanol instead
of clotted blood and the measurements of the sound speed
were carried out. The measured speed of sound in ethanol,
1200 m/ s at 20 C and 1150 m/ s at 37 C, agree well with
the tabulated values of 1191 m/ s at 20 C and 1159 m/ s at
37 C CRC, 1979.
D. Frequency-dependent attenuation
We employed an insertion loss method to measure the
ultrasound attenuation in clotted blood Hill et al., 2004.
The measurements of the pressure amplitude attenuation co-
efcient were conducted in a Lucite water tank with di-
mensions of 685555 cm
3
lled with degassed and
deionized water at 222 C for three center frequencies:
120 kHz, 1 MHz, and 3.5 MHz. The sample holder, shown
in Fig. 3, consisted of a Lucite frame which supported two
acoustic windows Tegaderm, 3M Health Care, St. Paul,
MN sandwiching the clot sample in saline solution. Each
clot was placed in the focal plane of one of the calibrated
ultrasonic transducers. Two holders were employed in our
experiments. The rst had an opening diameter of 70 mm
and was used for the 120-kHz experiments. The second had
an opening diameter of 40 mm and was used for experiments
at 1 and 3.5 MHz. In both cases the diameter of the openings
was much bigger than the acoustic wavelength and the trans-
ducer beamwidth at the corresponding frequencies. No dis-
tortion of acoustic eld due to diffraction of ultrasound was
evident in the transverse beam proles obtained when the
clot sample was absent.
All transducers were preliminarily calibrated in a water
tank at room temperature 201 C. Their experimentally
measured focal distance, focal depth, and beam width are
presented in Table I. The focal distance and beamwidth for
the 120-kHz transducer were 74 and 22 mm, respectively.
The 1-MHz transducer focal distance was 94 mm and the
beamwidth was 7 mm. The focal length for the 3.5-MHz
transducer was 93 mm and the beamwidth was 2.5 mm. The
transducers were driven by signals from a function generator
Agilent 33250 as shown in Fig. 3. For the experiments at
120 kHz, an Ultra Series 2021LF/HF linear amplier T&C
Power Conversion Inc., Rochester, NY with an accompany-
ing 50- impedance matching unit Sonic Concepts, Inc.,
Woodinville, WA amplied the driving signal. For the ex-
periments at 1 and 3.5 MHz, an AR 150LA Ampier Re-
search, Souderton, PA with no additional impedance match-
ing was utilized. At each frequency the pulse repetition
period was 1 ms corresponding to a 1-kHz PRF and each
tone burst contained 10 cycles. A TC4038 hydrophone RE-
SON, Inc., Goleta, CA and a 0.2-mm needle hydrophone
HP 0.2-mm interchangeable probe, Precision Acoustics Ltd.,
Dorset, UK were used to measure the peak-to-peak pressure
amplitude, initially in the free eld without the clot and
subsequently with the clot in the transducers focal plane.
The position of the hydrophones in the water tank was com-
puter controlled by a Velmex system with a stepping motor
controller NF 90 series, Velmex, Inc., Bloomeld, NY. A
sheet of acoustic absorbing material Aptex F28, Precision
Acoustics Ltd., Dorset, UK was xed on the far wall of the
Plexiglas tank to prevent the reection of acoustic waves.
By comparing the acoustic pressure amplitude at the
same location in the free eld with that from the intervening
clot, the magnitude of the coefcient of attenuation was mea-
FIG. 3. Experimental apparatus for the measurements of the amplitude at-
tenuation coefcient in blood clots.
TABLE I. Characteristics of transducers and hydrophones used in the mea-
surements of ultrasound attenuation in blood clots.
Transducer
frequency Hydrophone
Transducer
aperture
cm
3 dB
beamwidth
cm
Focal
distance
cm
120 kHz TC4038,
RESON, Inc.,
Goleta, CA
6.14 2.2 7.4
1 MHz Precision
Acoustics 0.2
mm needle
probe
2.5 0.7 9.4
3.5 MHz Precision
Acoustics 0.2
mm needle
probe
1.9 0.25 9.3
J. Acoust. Soc. Am., Vol. 119, No. 6, June 2006 Nahirnyak et al.: Physical properties of clotted blood 3769
sured. The pressure amplitude coefcient of attenuation, ,
was determined using the following formula Leighton,
1997:
=
1
l
ln
P
free
x, y
P
clot
x, y
, 2
where l is the length of clot along the ultrasound beam axis,
P
free
x, y is the pressure amplitude in the focal plane of
ultrasound transducer, P
clot
x, y is the pressure amplitude
at the same location after the clot was introduced into the
acoustic field, and x and y are the transverse coordinates
of the position of the hydrophone in the focal plane of
transducer Fig. 3. In our experiments, the thickness of
clots ranged from 0.69 to 0.89 mm. The thickness was as-
sessed with a calibrated caliper before placing the clot
into the holder. The magnitude of the pressure amplitude
was measured over the transverse beam profile using a
temporal peak, spatial peak pressure output of 0.18 MPa
for all three transducers. The transverse beam profile of
the 120-kHz transducer with the presence and absence of
the clot is shown in Fig. 4. The pressure amplitude attenu-
ation coefficient for each frequency was determined by
averaging the pressure reduction data within three blood
clot samples by scanning the transverse beam profile
along two perpendicular directions. The results of these
measurements were averaged and the mean and the stan-
dard deviation were computed to assess the variability.
III. RESULTS
The clot densities were 1.080.02 10
3
kg/ m
3
for
human clots and 1.060.01 10
3
kg/ m
3
for porcine clots.
The speed of sound in porcine clots was 15471 m/ s at
20 C and 15772 m/ s at 37 C. The speed of sound in
human clots was 15979 m/ s at 20 C and 16334 m/ s at
37 C.
The results for the measurements of amplitude attenua-
tion coefcients in human and porcine clots with the corre-
sponding standard deviations are presented in Table II. Their
magnitudes vary between 0.10 and 0.3 Np/ cm in porcine
clots and between 0.09 and 0.23 Np/ cm in human clots over
the frequency range between 120 kHz and 3.5 MHz. A least
squares t to the acoustic attenuation data in both porcine
and human clots shown in Fig. 5 demonstrates that the fre-
quency dependence of the amplitude attenuation coefcient
obeys approximately a one-third power law.
The average magnitudes of the coefcients of specic
heat of the human and porcine clots were 3.50.8
10
3
J/ kg K and 3.20.5 10
3
J/ kg K, respectively.
The coefcient of thermal conductivity for porcine clots
was 0.55 0.13 W / m K and for human clots was
0.590.11 W/ m K.
IV. DISCUSSION
An analysis of the results obtained should start with a
look at the structure of clotted blood presented in Fig. 6. A
formed blood clot has the same main constituents as whole
blood: red cells or erythrocytes, platelets, and brin bers.
Density, specic heat, and thermal conductivity for whole
human blood can be found in the literature. The density re-
ported to be 1.0510
3
to 1.610
3
kg/ m
3
, the specic heat
is 3.6110
3
to 3.8910
3
J/ kg K, and the thermal conduc-
tivity 0.49 to 0.55 W/ m K NCRP, 1992; Duck, 1990; Leh-
mann and DeLateur, 1982. Human red blood cells have a
density of 1.09310
3
kg/ m
3
Duck, 1990. Their specic
heat is 3.2110
3
J/ kg K and thermal conductivity is
0.450.482 W/ m K Toubal et al., 1999; Duck, 1990. Iso-
lated brin bers have density of 1.310
3
kg/ m
3
Mourad
and Kargl, 2003. The normal value for the human blood
hematocrit lies in the range of 39% to 52% Alanen et al.,
FIG. 4. Beam proles in the focal plane of the 120
-kHz transducer. The decrease of the pressure amplitude
lower curve was caused by the introduction of a clot
sample into the acoustic eld. The relative reduction of
the acoustic pressure at the center of the beam was used
to determine the amplitude coefcient of attenuation.
Error bars represent the standard deviation of the mea-
surements in three clots. The Tegaderm lms used in
the experiments do not cause a signicant change of the
pressure amplitude due to attenuation in the material of
the lm.
TABLE II. Pressure amplitude attenuation coefcients in whole porcine and
human blood clots.
Frequency
Attenuation in
clotted porcine
blood
Np/cm
Attenuation in
clotted human
blood
Np/cm
120 kHz 0.100.01 0.090.02
1 MHz 0.170.02 0.170.02
3.5 MHz 0.30.1 0.230.04
3770 J. Acoust. Soc. Am., Vol. 119, No. 6, June 2006 Nahirnyak et al.: Physical properties of clotted blood
1985. For weanling age 6 to 10 weeks pig blood, the
value of hematocrit ranges between 29% and 33% Osborne
and Meredith 1971 and is slightly anemic.
It is helpful to compare the measured acousto-
mechanical and thermal properties of clotted blood with
other types of soft tissues. Typical values for the density of
the human organs such as kidney, liver, and brain range be-
tween 1.0410
3
and 1.0610
3
kg/ m
3
Lehmann and De
Lateur 1982. Thus, the density of blood clots, 1.080.02
10
3
kg/ m
3
for human clots and 1.060.01
10
3
kg/ cm
3
for porcine clots, is consistent with the upper
range of values for soft tissue densities. The coefcient of
specic heat for brain is 3.6810
3
J/ kg K, for liver is
3.6010
3
J/ kg K, and for kidney is 3.8910
3
J/ kg K Le-
hmann and De Lateur, 1982. The thermal conductivity of
brain is 0.528 W/ m K, that of liver is 0.565 W/ m K, and
that of kidney is 0.544 W/ m K Lehmann and De Lateur
1982. Note that our data for the thermal properties of clotted
blood are also within the range of values for soft tissues in
the literature.
Our results of the measurements of sound speed in clot-
ted blood compare favorably with previously published re-
sults Grybauskas et al., 1978; Shung et al., 1984. Grybaus-
kas et al. 1978 obtained an experimental value for sound
speed in human blood clots of 1602 m/ s that were 90 min
old at 37 C. Shung et al. 1984 measured a sound speed of
1590 m/ s for human blood clots that were 1 h old at 23 C.
These investigators also noted an increase of sound speed
with the age of the clot. Our measured value for the speed
sound at 20 C, 15979 m/ s at 20 C, agrees with these
data.
Previously published data for the attenuation of ultra-
sound at 1 MHz in whole human blood are 0.017 Np/ cm
Hoskins et al., 2003. Grybauskas et al. 1978 reports
1.221.38 Np/ cm for the magnitude of absorption of ultra-
sound in the frequency range between 4.5 and 4.8 MHz.
Shung et al. 1984 obtained 0.45 Np/ cm for ultrasound at-
tenuation in human clotted blood at 7.5 MHz. In our experi-
ments, the attenuation of ultrasound in porcine and human
clotted blood at 1 MHz was 0.17 Np/ cm. If we compare the
obtained results with the published ones, we see that the
values obtained by Grybauskas et al. 1978 are signicantly
higher than the typical attenuation of ultrasound in soft tis-
sues for this frequency range. These data are closer to the
ultrasound attenuation in bone. On the other hand, our ex-
perimental results are in good agreement with the data pre-
sented by Shung et al. 1984.
The measured attenuation coefcients agree well with
the published values for other types of soft biological tissues
such as liver, kidney, and brain, which contain the same
types of proteins. For example, the value of ultrasonic at-
tenuation in liver is 0.17 Np/ cm, in muscles is
0.20.25 Np/ cm, and in brain is 0.12 Np/ cm Damask,
1978. All of them have the same order of magnitude of the
amplitude attenuation coefcient. The attenuation of the ul-
trasound in clotted blood is inuenced by ultrasonic absorp-
tion by macromolecules of the various proteins which con-
stitute clot NCRP, 1992.
A two-parameter least squares t was performed on the
attenuation data of the form =af
b
, where is the pressure
amplitude attenuation coefcient and f is the center fre-
quency. The power dependence, b, was 0.32 for porcine clots
and 0.28 for human clots. In contrast to the linear depen-
dence of attenuation on frequency, which is typical in many
soft tissues, our results demonstrated an approximate one-
third power dependence between 120 kHz and 3.5 MHz. The
frequency dependence of attenuation in clotted blood is simi-
lar to the frequency dependence of ultrasound attenuation in
lung 0.2 power, skeletal muscles 0.8 power, and skin
0.6 power Hill et al., 2004. Theoretical calculations of
the frequency dependence of the acoustic absorption in
whole clotted blood were reported by Mourad and Kargl
2003 and exhibited a 0.9 power dependence.
The standard deviation in the thermal conductivity and
specic heat data reects the natural variability of the physi-
cal properties of blood clots. The presence of air pockets,
cavities, and the inhomogeneous brin structure in clots
might also contribute to the standard deviation of the thermal
conductivity and specic heat experimental results. The
simple techniques used in this work may easily be employed
FIG. 5. Frequency dependence of the pressure amplitude attenuation coef-
cient in blood clots. The solid and dashed lines refers to a least square t of
the data to the form =a f
b
, where is the pressure amplitude attenuation
coefcient, f is the center frequency, and the R values listed in the legend
are the correlation coefcients and represent a goodness of t.
FIG. 6. Photo of the cross section of a porcine clot, with 20 magnication
stained with an antibrinogen antibody Dako Corp., Carpinteria, CA
which stains brin bers brown Datta, 2005.
J. Acoust. Soc. Am., Vol. 119, No. 6, June 2006 Nahirnyak et al.: Physical properties of clotted blood 3771
to investigate other types of biological soft tissues.
V. CONCLUSIONS
The acousto-mechanical and thermal properties of the
human and porcine clotted blood were determined experi-
mentally and add to the knowledge base of the physical
properties of clotted blood. The results obtained will allow us
to make numerical estimates of a magnitude of the thermal
elevation during ultrasound insonication of blood clots for
enhanced thrombolysis and may be helpful in the develop-
ment of high-intensity focused ultrasound and other thera-
peutic ultrasound applications.
Understanding the potential for thermal mechanisms in-
volved in the interaction of ultrasound with blood thrombi
during ultrasound-assisted thrombolysis is an important rst
step in improving thrombolytic efcacy while minimizing
unwanted thermal bioeffects.
ACKNOWLEDGMENTS
The authors want to thank Sampada Vaidya, Jun Tan,
Ashima Dhamija, and Nazli Bavani for the help in prepara-
tion of the blood clot samples. This research was supported
by the National Institutes of Health, Grant No. NIH/NINDS
R01-NS047603.
Alanen, A., and Kormano, M., 1985. Correlation of the ecogenisity and
structure of clotted blood, J. Ultrasound Med. 4, 421425.
American Heart Association AHA 2005. Heart Disease and Stroke
Statistics2005 Update., American Heart Association, Dallas, TX.
Blinc, A., Francis, C. W., Trudnovski, J. L., and Carstensen, E. L. 1993.
Characterization of ultrasound-potentiated brinolysis in vitro, Blood
81, 26362643.
Cintas, P., Nguen, F., Boneu, B., and Larrue, V. 2004. Enchancement of
enzymatic brinolysis with 2-MHz ultrasound and microbubbles, J.
Thrombosis Haemostasis 2, 11631166.
CRC 1979. CRC Handbook of Chemistry and Physics edited by Robert C.
West, CRC, Baca Ratan, FL.
Damask, A. C. 1978. Medical Physics. Volume I: Physiological Physics,
External Probes Academic, New York, p. 215.
Datta, S., McAdory, L. E., Tan, and J., Holland, C. K. 2005. Cavitation
detection during ultrasound-assisted thrombolysis in porcine blood clots,
J. Acoust. Soc. Am. 117, 2558.
Dick, A., Neuerburg, J., Schmitz-Rode, T., Alliger, H., Schmid-Schonbein,
H., and Gunter, R. W. 1998. Thrombolysis of mural thrombus by ultra-
sound. An experimental in vitro study, Invest. Radiol. 33, 8590.
Duck, F. A. 1990. Physical Properties of Tissue: A Comprehensive Refer-
ence Book Academic, London.
Everbach, E. C., and Francis, C. W. 2000. Cavitational mechanisms in
ultrasound-accelerated thrombolysis at 1 MHz, Ultrasound Med. Biol.
267, 11531160.
Francis, C. W. 2001. Ultrasound-enhanced thrombolysis, Echocardiogr.
183, 239246.
Francis, C. W., Blinc, A., Lee, S., and Cox, C. 1995. Ultrasound accel-
erates transport of recombinant tissue plasminogen activator into clots,
Ultrasound Med. Biol. 213, 419424.
Francis, C. W., Onundarson, P. T., Carstensen, E. L., Blinc, A., Meltzer, R.
S., Schwarz, K., and Marder, V. 1992. Enhancement of brinolysis in
vitro by ultrasound, J. Clin. Invest. 90, 20632068.
Grybauskas, P., Kundrotas, K., Sukackas, V., and Jaronis, E. 1978. Ultra-
sonic digital interferometer for investigation of blood coagulation, Ultra-
sonics 161, 3336.
Harpaz, D., Chen, X., Francis, C. W., Marder, V. J., and Meltzer, R. S.
1993. Ultrasound enhancement of thrombolysis and reperfusion in
vitro, J. Am. Coll. Cardiol. 21, 15071511.
Hill, C. R., Bamber, J. C., and ter Haar, G. R. 2004. Physical Principles of
Medical Ultrasonics Wiley, Chichester, England, p. 123.
Holland, C. K., Vaidya, S. S., Coussios, C.-C., and Shaw, G. J. 2002.
Thrombolytic effects of 120 MHz and 1 MHz ultrasound and tissue plas-
minogen activator on porcine whole blood clots, J. Acoust. Soc. Am.
112, 2370.
Hoskins, P. R., Trush, A., Martin, K., and Whittingham, T. A. 2003. Di-
agnostic Ultrasound. Physics and Equipment GMM, London.
Kreith, F., and Black, W. Z. 1980. Basic Heat Transfer Harper and Row,
New York.
Lauer, C. G., Burge, R., Tang, D. B., Bass, B. G., Gomez, E. R., and Alving,
B. M. 1992. Effect of ultrasound on tissue-type plasminogen activator-
induced thrombolysis, Circulation 86, 12571264.
Lehmann, J. F., and DeLateur, B. J. 1982. Therapeutic Heat and Cold
Williams & Wilkins, Baltimore.
Leighton, T. G. 1997. The Acoustic Bubble Academic, London, p. 28.
Mourad, P. D., and Kargl, S. G. 2003. Acoustic properties of uid-
saturated blood clots, Technical Report of the Applied Physics Labora-
tory, University of Washington. APL-UW TR 2003, p. 16.
National Council on Radiation Protection and Measurements NCRP
1992. Exposure criteria for medical diagnostic ultrasound: I. Cretiria
based on thermal mechanisms, NCRP Publications, Report no. 113, June
1, p. 35.
Olsson, S. B., Johansson, B., Nilsson, A. M., and Olsson, C., Roijer, A.
1994. Enhancement of thrombolysis by ultrasound, Ultrasound Med.
Biol. 20, 375382.
Osborne, J. C., and Meredith, J. H. 1971. Hematological values of the
normal weanling piglet, Cornell Vet. 611, 1323.
Polak, J. F. 2004. Ultrasound energy and the dissolution of thrombus, N.
Engl. J. Med. 35121, 21542155.
Sakharov, D. V., Hekkenberg, R. T., and Rijken, D. C. 2000. Acceleration
of brinolysis by high-frequency ultrasound, Thromb. Res. 1004, 333
340.
Semat, H. 1953 Fundamentals of physics. Rinehart, New York, p. 275.
Shaw, G. J., Bavani, N., Dhamija, A., and Lindsell, C. J. 2006. Effect of
mild hypothermia on the thrombolytic efcacy of 120 kHz enchanced
thrombolysis in an in-vitro human clot model, Thromb. Res. 117, 603
608.
Shaw, G. J., Hahn, N. L., Wagner, K. R., Kanter, D. S., and Holland, C. K.
2001a. Ultrasound assisted clot lysis for stroke therapy, J. Acoust. Soc.
Am. 109, 2456.
Shaw, G. J., Hahn, N. L., Wagner, K. R., Kanter, D. S., and Holland, C. K.
2001b. Ultrasound assisted thrombolysis in an in vitro clot model,
Acad. Emerg. Med. 8, 542.
Shung, K. K., Fei, D. Y., Yuan, Y. W., and Reeves, W. C. 1984. Ultrasonic
characterization of blood during coagulation, J. Clin. Ultrasound 123,
147153.
Siddiqi, F., Blinc, A., Braaten, J., and Francis, C. W. 1995. Ultrasound
increases ow through brin gels, Thromb. Haemostasis 733, 495498.
Suchkova, V., Carstensen, E. L., and Francis, C. W. 2002. Ultrasound
enhancement of brinolysis at frequencies of 27 to 100 kHz, Ultrasound
Med. Biol. 283, 377382.
Toubal, M., Asmani, M., Radziszewski, E., and Nongaillard, B. 1999.
Acoustic measurement of compressibility and thermal expansion coef-
cient of erythrocytes, Phys. Med. Biol. 445, 12771287.
Wolpert, S. M., Bruckmann, H., Greenlee, R., Wechsler, L., Pessin, M. S.,
and del Zoppo, G. J. 1993. Neuroradiologic evaluation of patients with
acute stroke treated with recombinant tissue plasminogen activator,
AJNR Am. J. Neuroradiol. 141, 313.
3772 J. Acoust. Soc. Am., Vol. 119, No. 6, June 2006 Nahirnyak et al.: Physical properties of clotted blood